[ctdb]CTDB CRASH

[Christian Ambach]

Hi,

On 01/19/2012 09:25 AM, guoleia at 139.com wrote:
> Hi folks,
>
> I'm using ctdb for a samba cluster.
> ctdb version:1.0.114-1
> 1¡¢There're running cluster(including node_A,node_B,node_C), the master primary node is node_A.
> 2¡¢Adding a new node(node_D), and append the ip of node_D to the "/etc/ctdb/nodes" file in node_A and node_B and node_C,
> make the file which's in the catalog "/etc/ctdb/nodes" of node_D consistent with the others.
> 3¡¢Run the command "ctdb reloadnodes" on node_A, meanwhile running the command "service ctdb start" on node_D , finding that the operation will cause ctdb process crash in some nodes.
>
> ***************************************************************
> Program terminated with signal 6, Aborted.
> #5<signal handler called>
> #6  0x0000000000445b62 in ctdb_canonicalize_ip (ip=0x7348a00000002, cip=0x7fff0deeab90) at common/ctdb_util.c:504
> #7  0x0000000000445c3f in ctdb_same_ip (tip1=0x7348a00000002, tip2=0x2b5d9d4a7fc0) at common/ctdb_util.c:522

Do you have blank lines in the /etc/ctdb/nodes file?
You must also have all servers in the same order on all of the nodes.

Cheers,
Christian